A semi-empirical approach to estimate vertical transport by nonprecipitating convective clouds on a regional scale

A semi-empirical approach to estimate vertical transport by nonprecipitating convective clouds on a regional scale
of 16
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Atmospheric Environment Vol. 24A, No. 8, pp. 2153 2168, 1990. 0004 6981/90 $3.00+0.00 Printed in Great Britain. Pergamon Press plc A SEMI-EMPIRICAL APPROACH TO ESTIMATE VERTICAL TRANSPORT BY NONPRECIPITATING CONVECTIVE CLOUDS ON A REGIONAL SCALE FRED M. VUKOVICH Atmospheric and Marine Program, Research Triangle Institute, P.O. Box 12194, Research Triangle Park, NC 27709, U.S.A. and JASON K. S. CHING* Atmospheric Sciences Modeling Division, Atmospheric Research and Exposure Assessment Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711, U.S.A. (First received 19 April 1989 and in finalJbrm 29 December 1989) Abstract A semi-empirical approach to estimate the vertical flux of mass between the boundary layer and the cloud layer by an ensemble of nonprecipitating cumulus clouds has been developed. The model determines the existence of the cloud ensemble, estimates the cloud amount at cloud base, and establishes he vertical distribution of the convective cloud amount attributed to a cloud population, having a continuous spectrum of cloud depth, using standard meteorological data. The mass flux is then estimated for the ensemble or for a 'processor' cloud, a single cloud which, on the average, can be used to represent the ensemble, using a modified version of the cloud model developed by Ritter and Stedman (1985, Coop. Agreement No. CR 807485-01, 02). Results of a sensitivity analysis are presented. The analysis examined the behavior of the model as the vertical distribution of temperature and dew point changed from one atmospheric state to another. Key word index: Nonprecipitating cumulus clouds, cloud flux model, vertical transport. 1. INTRODUCTION or can undergo chemical transition through liquid phase chemistry which occurs in the cloud into which Cumulus clouds have their roots within the planetary they were initially entrained. The details of the trans- boundary layer (PBL). Warm moist air in convective port processes, as well as chemical processes, associ- eddies cools in rising thermals. At the lifting condensa- ated with cumulus convection are indeed a complex tion level (LCL), the water vapor in the thermals problem. condenses and the parcel gains additional buoyancy A significant body of information exists on the due to the release of latent heat of condensation. The influence of large-scale convective storm systems on parcel, now a cloud, will rise to greater altitudes until the redistribution of mass and heat (Deardorff, 1975; the additional buoyancy is spent. The noncloud air in Betts, 1976: Seguin and Garstang, 1976; Johnson, the free troposphere will subside in a complex manner 1976; Riehl and Malkus, 1958, 1961; Ooyama, 1964; to maintain mass balance. The rising air parcel within Yanai, 1964; Charney and Eliassen, 1964). Very little is the clouds may contain a variety of pollutants from known about the influence of a large-scale system of the boundary layer (see Fig. 1). When a nonprecipitat- nonprecipitating convective clouds on the redistribu- ing cloud with this pollutant burden dissipates, the tion process, though recent research results are pro- transport and diffusion of pollutants are then con- viding a background focusing on the importance of strained to the free-tropospheric flow regime, and, this effect (Chatfield, 1982; Gidel, 1983; Vukovich et subsequently, the pollutants undergo horizontal dis- al., 1985; Ching, 1986; Greenhut, 1986). Widespread placements in directions that generally differ from that systems of nonprecipitating convective clouds often in the underlying boundary layer (Ching et al., 1988). develop, particularly in the summertime, in moist They may eventually reenter the boundary layer high-pressure systems. GOES visible and infrared through precipitation systems or by re-entrainment imagery have detected these systems, over the contin- through mixing from below, or they may remain in the ental U.S., having areas greater than one million free troposphere where they will eventually dissipate square kilometers. These systems are almost daily phenomena in the summertime and are generally found diurnally in the period from 11 a.m. to 7 p.m. * On assignment from National Oceanic and Atmospheric local time (or roughly for 8 h per day). Initiation of the Administration, United States Department of Commerce. system takes place with the break-up of the surface- 2153  2154 FRED M. VUKOVICH and JASON K. S. CHING Pollution Remnont of , ; Poll ted Bounda.ry Layer Fig. 1. Hypothetical structures of air pollution and convective clouds in the lower tropo- sphere; shaded area represents air pollution. based inversion formed during the night. The surface- must be accounted for in regional-scale air quality based inversion usually breaks up a few hours before models in order to properly assess and predict air the surface heat flux is a maximum. Dissipation of the quality on that scale. A cloud flux model, CUVENT, system generally begins with the reformation of the was developed in order to estimate the effect of surface-based inversion in the late afternoon and transport of pollutants out of the boundary layer into evening. For the most part, the system is made up of the free troposphere by an ensemble of nonprecipitat- nonprecipitating convective clouds whose vertical ex- ing, subgrid-scale convective clouds. This paper de- tent varies from 0.5 to 2.5 km above cloud base, and scribes that model and the methodology used to whose average vertical extent is about 1.5 km above develop the various parameterization schemes for cloud base, about equal to the average vertical extent closure of the model• of the planetary boundary layer (PBL). Over an 8-h Before discussing the model, the distinction between period, a single or a small cluster of convective clouds cloud layer and the PBL as it is used in the context of would have little effect on the boundary layer mass this paper must be explained. The cloud layer is that distribution, but the large scale convective system layer immediately above the top of the PBL in which cloud 'vent' a large fraction of the PBL over an venting nonprecipitating convective clouds develop enormous area, suggesting a process that can have a (see Fig. 2) in the large-scale convective system in significant effect on the pollution burden in the PBL. question. Cloud base is generally found at or immedi- Since the transport processes that are associated ately below the top of the PBL. Theclouds in the cloud with the convection can have an impact on the layer are those that penetrate above the top of the pollutant burden in the mixed layer and in the free PBL; these clouds can transport mass, heat, and troposphere, and on the cloud chemistry, this process possibly momentum out of the PBL into the cloud  Vertical transport estimation 2155 ~~;J~~}~':.:.'~~iii~.i' ..' .~t?~'-~ i ',, k'tr ~''~ "~' " '~k~ ~'~':': :' J:: :t:'x ; ' """~':": ,~ff'z_'"';~ Fig. 2. Graphical depiction of cloud layer and boundary layer. zi I - -- ~ Z LCL 0 ~ZIN °.. ..'..:.:'.::.2.:':;.:?-'::.':~..::..::~.. ~ ... ".'.. "'.'; ; "' ':.':::".r~...::: "..:'-~4 • ..:..':::. '....~.',~ ~;' 2: .~",~.".,..i • • ,..'t m.,.,,, ....'~ : : :- .'.'.~ =. .'~:= := ' - -' ~', ~i ~ ...,~..;:;::::c...:-"" '...-:.::." i::"=.i:i~iii ~iii.:~i~i~ "'"" ' .:.:ii:: "..C :,~ii iiiiiiiiiiiiiliiiili iiiiiiiiiiiiiiii ~--'-:kiiiiiii ==i ~;iiiiii Fig. 3. Diurnal cycle of the top of the PBL(Z~), the height of the LCL (ZLcL) and the top of the surface based inversion (Z1N) based on the results from VENTEX. Dots depict zone of variability about Z~ and ZLC L. The cross-hatched zone is the time at which convective clouds exist. layer which is part of the free troposphere. The top of produced just before Zi = ZLC L because there is a zone the cloud layer can be generally defined, in the context of variability about the mean values Z i and ZLC L of this study, as the height of the top of the highest (represented by the dots about the lines that represent nonprecipitating convective cloud in the system. Lat- Z i and ZLC L in Fig. 3) due to differential processes. er, in this paper, the cloud layer thickness will be This aspect will be discussed in detail later• It suffices quantified, here to note that, at this time, the cloud amount is The cloud system which is the focus of this paper is small and all clouds are venting or active clouds since produced through processes that occur on a diurnal Zi~<ZLc L. cycle. After sunrise, the surface sensible heat flux After crossover and ZLC L< Zi(cross-hatched area in begins to dissipate the nocturnal surface-based inver- Fig. 3), the cloud system becomes a combination of sion layer (Fig. 3). When the convective process com- shallow clouds that do not penetrate through Z~ (i.e. pletely dissipates the nocturnal inversion, a rapid fair weather cumulus) and active clouds. The cloud growth of the PBL usually follows. Meanwhile, the amount reaches its peak value at this time and that mean height of the lifting condensation level relative value is a function of PBL parameters such as the to the surface, ZLc L, rises due to the increase in surface surface heat flux, PBL moisture content, ZLCL, Zi, etc. temperature which increases the surface dew point Late in the afternoon, the surface based inversion is re- depression, but it does not rise as rapidly as the established, the production of convective elements average top of the PBL, Zi (i.e. the top of the convec- ceases, and the cloud amount approaches zero. Some tive layer). When Zi=ZLcL, convective clouds are convective clouds usually remain; these are large produced. Actually, convective clouds are usually active clouds which have reached their level of free  2156 FRED M. VUKOVICH and JASON K. S. CHING convection and are, in a sense, self-perpetuating. These convective cloud amount (section 6), are determined. clouds generally dissipate in the evening. The vertical distribution of the convective cloud amount is a function of the cloud depth distribution for the convective cloud system. After these para- 2. VENTEX meters are determined, they are combined with the standard meteorology data in a cloud model discussed Most of the data used to develop parameters for in section 7. This model is based on the CUVENT were obtained as a part of a field program Arakawa-Schubert (1974) scheme, was developed by called VENTEX (VENTing EXperiment). The study Ritter and Stedman (1985), and modified using pro- took place in the rural regions outside of Lexington, cedures presented by Johnson (1977) to estimate cloud Kentucky, in the period 9 July-16 August 1984. The dynamic and thermodynamic parameters, in particu- field program was designed to obtain data to charac- lar the flux vertical velocity associated with the en- terize and study processes associated with convective semble of nonprecipitating convective clouds (section clouds. The main participants in the program were 8). Sensitivity analyses of this model are discussed in Argonne National Laboratory, Pacific Northwest La- section 9. boratory, the University of Kentucky, and the Re- search Triangle Institute. Measurements included sur- 4. TEST FOR NONPRECIPITATING CONVECTIVE CLOUDS face heat flux using the eddy correlation technique; subjective (observer) and objective (cloud photo- Four tests are performed for screening purposes to graphs) cloud amount observations; cloud height and determine the existence of nonprecipitating convective dimensions using a laser rangefinder and from aircraft clouds at an instant of time. The first test uses the data observations, profiles of temperature and humidity from the host model which specifies whether rain from RAOBs, Airsondes, and aircraft observations; occurs in the grid. If not, the second test is executed SODAR observations; and surface temperature, dew which establishes whether the season will support the point, and wind observations at various locations, production of the large convective systems. (N.B.: this GOES visible and infrared data were collected during test was primarily created to limit the call for CUV- the VEXTEX period, and these data were combined ENT, which was used as a subroutine, by the host with the surface-based cloud observations to obtain computer model to those periods when large convec- information on the characteristics of the cloud width tive systems were most likely to exist in order to frequency distribution as a function of cloud amount, reduce computer running time.) The third test deter- Some details of the VENTEX program have been mines if the surface heat flux is sufficient to produce published in a report by Vukovich and Haws (1987) convective thermals, and the fourth test determines if which is available through NTIS. moist convection is occurring. The large systems of nonprecipitating convective clouds that are accounted for by CUVENT have a 3. STRUCTURE OF THE METHOD definite warm season preference (i.e. late spring, sum- mer, and early fall over the U.S.). In order to examine CUVENT was developed for adaptation into re- this warm season preference and to determine restric- gional-scale air quality models, and was designed to be tion limits, a frequency analysis was performed that self-consistent; i.e. CUVENT would operate within matched the occurrence of nonprecipitating cumulus the framework of the information and data provided clouds with the surface dry bulb temperature and the by the host model and determine all the parameters surface dew point temperature. For the frequency needed to execute the calculation from the provided analysis, 3 years of data (1975-1977) at 3-h intervals data. Therefore, CUVENT accepts basic surface and from five National Weather Service synoptic weather upper air meteorological data (e.g. pressure, temperat- stations were used: Buffalo, New York; Louisville, ure, humidity, surface heat flux data, etc.), which is Kentucky; St. Louis, Missouri; Jackson, Mississippi; generally provided by most sophisticated air quality and Raleigh-Durham, North Carolina. The large sys- models, to develop model parameters such as the terns of nonprecipitating cumulus clouds have been convective cloud amount at cloud base, convective observed to have a diurnal behavior and are normally cloud distribution parameters, etc. Various empirical found during the daytime between 1100 and 1900 models that were developed using data from, for the EST. Therefore, the only usable data were obtained most part, the VENTEX program and theoretical from the meteorology data set at 1200, 1500 and 1800 procedures that were described in the literature, were EST. The statistical data base was limited to observa- used to provide closure for the algorithm, tions of cumulus clouds having cloud amounts greater Figure 4 provides a flow diagram for the CUVENT than or equal to 0.2. Thus, a total of 797 cases met the algorithm. The model initially determines whether criteria that nonprecipitating cumulus existed over the conditions allow nonpreeipitating cumulus clouds to 3 year period and these were distributed as follows: exist. This is discussed in section 4. Under favorable Jackson, Mississippi, 163; Raleigh-Durham, North conditions, the convective cloud amount at cloud base Carolina, 166; St. Louis, Missouri, 147; Buffalo, New (section 5) as well as the vertical distribution of the York 126; and Louisville, Kentucky, 195.  Vertical transport estimation 2157 Input Meteorology Test For Existence of Nonprecipitating Cumulus Clouds 1 Determine Convective Cloud Amount at Cloud Base I Establish Vertical Distribution of Cloud Amount f Calculate Cloud Dynamic | and Thermodynamic 1 arameters -~loud Fig. 4. Flow diagram for CUVENT. Cumulative frequency distributions of the existence surface temperature exceeds 10°C coupled with sur- of cumulus clouds were created using all data and were face dew point temperature greater than 0°C permits a function of surface dry bulb temperature (Fig. 5A) nonprecipitating cumulus clouds. On the average, and of surface dew point (Fig. 5B). Similar frequency midday surface temperatures of 10°C and dew point distributions were constructed using the data for each temperatures of 0°C seldom characterize the period station. The frequency distributions that were created when the large-scale systems of nonprecipitating con- using station data were compared with the frequency vectives clouds that are the focal point of this study distribution that were created using all data in order exist. to determine whether a criterion can be established If the seasonal surface temperature and dew point based on surface temperature and surface dew point criteria were satisfied, then the next test determined temperature that can be universally applied over the whether the sensible heat flux from the surface to the region or whether such a criterion must be established atmosphere was sufficient to produce convective ther- regionally. For the stations selected, there did not mals. This test simply requires the heat flux to be seem to be a substantial difference between surface positive. The last test establishes whether the convec- temperatures and surface dew point temperatures tive elements have reached their lifting condensation when the selection criterion was based on 75-85 level (LCL) (i.e. became saturated) and produced a percentile values. Surface temperatures and surface cloud. dew point temperatures varied from one station to The heat content varies from one convective ele- another, based on the above percentile criterion, by ment to another so that the extent of penetration of about 2-4°C. Based on this comparison, it was decided one convective element through the boundary layer to use the 85 percentile value from the cumulative relative to that for another will vary. This variation frequency distribution that was created using all data produces a region in the atmosphere near the average to define the restriction limit for cumulus clouds in the top of the boundary layer where the actual top of the warm season. The 85 percentile value corresponded to boundary layer is highly variable. This region is called a surface temperature of 10°C and a surface dew point the entrainment zone, discussed by Deardorff et al. temperature of 0°C. Thus, the atmosphere where (1980). The top of the entrainment zone marks the

Guiao de Leitura

Mar 7, 2018

Drawing Process

Mar 7, 2018
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks